Aims and Objectives:

To investigate whether changes in the levels of physical activity, self-reports on dietary intake and habits are associated with subsequent changes in the development of body fat during early adolescence.

Design:

A five year prospective cohort of adolescents in the Avon Longitudinal Study of Parents and Children (ALSPAC) study.

Methods:

We propose to create a prospective cohort using four waves of ALSPAC data (recorded at ages 10.6, 11.8, 13.9 and 15). This cohort will form the basis of analysis into the independent and interdependent effects of changes in physical activity (total activity and MVPA) between the ages of 11.8-13.9, and dietary intake and habits (dietary energy density; fruit and vegetable intake; kJ/day from liquids) between the ages of 10.6-13.9, on subsequent changes in body fat (changes in Fat Mass and Fat Mass Index) between the ages of 13.9-15, after accounting for potential effect modification across known genetic (presence of high risk allele on FTO gene) and socioeconomic risk factors for increases in body fat in this cohort.

Main outcome measures:

Change in Fat Mass (kg) and Fat Mass Index (kg fat/ height) measured by Dual-energy X-ray Absorptiometry between the ages of 13.9-15.

Main predictors:

Changes in levels of physical activity:

Change in total activity (average counts per min over full period of valid recording (worn for greater than than or equal to 10 hours for greater than than or equal to 3 days)) and daily amount of moderate-vigorous physical activity (MVPA) (average min per day with a total of greater than 3600 counts) between the ages of 11.8-13.9.

Measurement of changes in dietary intake and other dietary habits:

Dietary energy density (kJ/g) (DED) excluding drinks from 3-day unweighed diet diaries. Percentage of DED from fruit and vegetables. DED from Daily energy intake (kJ/day) from: milk, fruit juice, fruit squash and cordials, fizzy drinks, water and flavoured water, hot chocolate, teas, coffee and alcohol. Dietary habits: percentage of DED consumed at breakfast, lunch, evening meal/ time of energy consumption; size of portions; frequency of breakfast consumption. Measured between the ages of 10.6-13.9.

Covariates of changes in body fat during adolescence:

We will examine whether the effect of changes in levels of physical activity, dietary intake and dietary habits on body fat are influenced by various genetic (presence of the high risk FTO genotype (rs9939609: comparing A:A; A:T; T:T), developmental (changes in pubertal development (Tanner stage) between the ages of 10.6 - 13.9) and socioeconomic (parental socioeconomic status, level of maternal education, mother and partner BMI) risk factors.

Statistical analyses:

Following descriptive analyses, the main method of analysis will be linear regression. Models will be produced for the regression of changes in fat mass and fat mass index between the ages of 13.9 - 15, on changes in levels of physical activity (ages 11.8 - 13.9) and dietary intake and habits (ages 10.6 - 13.9), with adjustment for established genetic, developmental, and socioeconomic risk factors for excess weight gain during adolescence. Potential effect modification will be tested in these models by including interaction terms for: changes in physical activity (greater than than or equal to 15 minutes a day of MVPA to less than 15 minutes) x fruit and vegetable consumption (greater than than or equal to 5 portions a day to less than 5 portions a day) and dietary intake and physical activity variables with participant sex.

Expected outcomes:

The proposed study will provide the information necessary to more effectively identify the behavioural targets of obesity prevention programmes by, 1) modelling the independent and interdependent effects of changes in dietary intake, dietary habits and levels of physical activity on changes in body fat, 2) providing a resource which will allow more precise estimates of the feasibility, likely effect and cost effectiveness of an intervention to be made, 3) evaluating the impact of naturally occurring changes to behaviour, within a context of known risk factors for obesity in adolescence. This method might also provide an empirical template for modelling behaviour change with observational data which can be applied in other behaviours or populations.